INDEX

Explanations

words related to being puzzled or confused

oai_token-act-pair · gpt-3.5-turbo

words that convey confusion, frustration, or emotional intensity

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 appropriated

-0.78

 faire

-0.78

ascript

-0.77

 screens

-0.74

otta

-0.72

cribed

-0.71

 stabilized

-0.67

 designated

-0.66

vertising

-0.65

database

-0.64

POSITIVE LOGITS

ness

0.91

ingly

0.89

ly

0.85

LY

0.82

 Beir

0.81

NESS

0.79

 Tact

0.73

nesses

0.71

 Dragonbound

0.70

 Puzz

0.68

Activations Density 0.124%

words related to being puzzled or confused

words that convey confusion, frustration, or emotional intensity

No Comments

No Known Activations